Stabilize hyperspherical UKF arbitrary-noise weights

[FlorianPfaff]

Summary

• normalize arbitrary-noise weights after scaling by their maximum finite value
• prevent overflow when individually finite weights have a non-finite direct sum
• reject empty arbitrary-noise supports with a clear boundary error
• add focused regressions for maximum finite weights and empty support

Bug

HypersphericalUKF.predict_nonlinear_arbitrary_noise() normalized weights with

noise_weights / noise_weights.sum()

For valid positive finite weights near np.finfo(float).max, the sum overflowed to infinity. The normalized weights collapsed to zero; the subsequent Cartesian-product weights were divided by zero and the predicted mean/covariance became non-finite.

Fix

First divide all positive weights by their maximum. The scaled weights lie in (0, 1], so their sum remains finite for practical sample counts. Dividing by that scaled sum gives the same mathematical normalized weights without overflow.

The method now also rejects an empty weight vector before attempting normalization.

Validation

• isolated numerical reproduction confirms the old direct sum overflows for two maximum finite weights
• regression runs the prediction under np.errstate(over="raise", invalid="raise", divide="raise") and verifies the expected unit mean and zero covariance
• regression verifies an empty noise support raises a parameter-specific ValueError
• branch comparison against main: two commits ahead, zero behind; implementation diff is three added lines

[github-actions]

✅MegaLinter analysis: Success

Descriptor	Linter	Files	Fixed	Errors	Warnings	Elapsed time
✅ COPYPASTE	jscpd	yes		no	no	21.68s
✅ JSON	prettier	7	0	0	0	1.47s
✅ JSON	v8r	7		0	0	3.58s
✅ MARKDOWN	markdownlint	68	0	0	0	1.95s
✅ MARKDOWN	markdown-table-formatter	68	0	0	0	0.81s
✅ PYTHON	black	1489	182	0	0	88.33s
✅ PYTHON	isort	1489	327	0	0	2.53s
✅ REPOSITORY	betterleaks	yes		no	no	2.49s
✅ REPOSITORY	checkov	yes		no	no	52.11s
✅ REPOSITORY	gitleaks	yes		no	no	14.19s
✅ REPOSITORY	git_diff	yes		no	no	0.35s
✅ REPOSITORY	secretlint	yes		no	no	58.97s
✅ REPOSITORY	syft	yes		no	no	5.32s
✅ REPOSITORY	trivy-sbom	yes		no	no	5.98s
✅ REPOSITORY	trufflehog	yes		no	no	31.36s
✅ YAML	prettier	11	0	0	0	0.72s
✅ YAML	v8r	11		0	0	11.31s
✅ YAML	yamllint	11		0	0	0.59s

Notices

📣 MegaLinter 9.5.0 is out! Discover the new features and security recommendations in the release announcement. (Skip this info by defining SECURITY_SUGGESTIONS: false)

See detailed reports in MegaLinter artifacts

Your project could benefit from a custom flavor, which would allow you to run only the linters you need, and thus improve runtime performances. (Skip this info by defining FLAVOR_SUGGESTIONS: false)

• Documentation: Custom Flavors
• Command: npx mega-linter-runner@9.6.0 --custom-flavor-setup --custom-flavor-linters PYTHON_BLACK,PYTHON_ISORT,COPYPASTE_JSCPD,JSON_V8R,JSON_PRETTIER,MARKDOWN_MARKDOWNLINT,MARKDOWN_MARKDOWN_TABLE_FORMATTER,REPOSITORY_CHECKOV,REPOSITORY_GIT_DIFF,REPOSITORY_GITLEAKS,REPOSITORY_BETTERLEAKS,REPOSITORY_SECRETLINT,REPOSITORY_SYFT,REPOSITORY_TRIVY_SBOM,REPOSITORY_TRUFFLEHOG,YAML_PRETTIER,YAML_YAMLLINT,YAML_V8R

Show us your support by starring ⭐ the repository